Speakeasy AI Control Plane: Securing Enterprise AI Adoption

Speakeasy CEO Sagar Batchu explains how their AI control plane helps organizations adopt AI broadly while maintaining governance. This system places every agent, tool, and MCP server behind a single security layer. It authenticates actions, enforces policies, and inspects sessions for prompt injection and data exfiltration. What's interesting is that George Finney, CISO at The University of Texas System, and Nick Espinosa, host of The Deep Dive Radio Show, are pressure-testing this approach. Nico Waisman, CISO at XBOW, notes that AI-generated code introduces vulnerabilities because models are trained on open-source repositories already full of them. Andrea Sivieri from CoreView points out that many Microsoft 365 security tools overlook configuration and permissions. The bottom line is that new solutions are emerging to address the complex security challenges presented by widespread AI adoption and expanding attack surfaces.